| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Gastrointestinal Imaging |
1 From the Department of Radiology and the Institute of Radiation Medicine, Seoul National University College of Medicine, 28 Yongon-Dong, Chongno-Gu, Seoul 110-744, Korea; and Clinical Research Institute, Seoul National University Hospital, Korea. From the 1999 RSNA scientific assembly. Received June 21, 2000; revision requested July 27; revision received September 26; accepted October 11. Address correspondence to J.K.H. (e-mail: hanjk@radcom.snu.ac.kr).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: The study group consisted of 109 hepatic hemangiomas in 69 patients who underwent two-phase spiral CT during 1 year. CT scans were obtained during the hepatic arterial (30-second delay) and portal venous (65-second delay) phases after injection of 120 mL of contrast material (3 mL/sec). Arterioportal shunts were diagnosed when hepatic arterial phase CT scans showed a wedge-shaped or irregularly shaped homogeneous enhancement in the liver parenchyma adjacent to the tumor and when portal venous phase CT scans showed isoattenuation or slight hyperattenuation, compared with normal liver in that area, and when there was no demonstrable cause of these attenuation differences. The presence of arterioportal shunt in hemangioma was correlated with the size of the tumor and the rapidity of intratumoral enhancement.
RESULTS: Arterioportal shunt was found in 28 (25.7%) of 109 hemangiomas. There was no statistically significant relationship between lesion size and presence of the arterioportal shunt (P = .653). Arterioportal shunt was more frequently found in hemangiomas with rapid enhancement (P < .01).
CONCLUSION: Arterioportal shunts are not uncommonly seen in hepatic hemangiomas at two-phase spiral CT. Hemangiomas with arterioportal shunts tend to show rapid enhancement.
Index terms: Angioma, gastrointestinal tract, 761.3194 • Computed tomography (CT), helical, 761.12112, 761.12114, 761.12115 • Computed tomography, phase imaging • Liver, CT, 761.12112, 761.12114, 761.12115 • Liver, MR, 761.121411, 761.12143 • Liver neoplasms, blood supply, 761.30 • Shunts, arterioportal, 952.453, 957.453
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
An arterioportal shunt associated with a hepatic tumor is generally recognized to be characteristic of malignant tumors (5) but is believed to be rare in hemangiomas (5–10). However, With the increased use of multiphasic spiral CT in the liver, more hemangiomas are imaged during the hepatic arterial phase (HAP), and atypical features for hemangiomas are more frequently seen. In our experience, a wedge-shaped or irregularly shaped parenchymal enhancement adjacent to the hemangioma was occasionally found at CT during the HAP, and this area usually changed to isoattenuation or slight hyperattenuation compared with normal liver during the portal venous phase (PVP). We believe that this finding represents the existence of an arterioportal shunt associated with the hemangioma. Therefore, we performed this study to determine the prevalence of arterioportal shunts associated with hepatic hemangiomas and to describe the two-phase spiral CT findings. We also correlated the presence of arterioportal shunt with the size and rapidity of enhancement of hemangiomas.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Diagnostic criteria of hemangioma were as follows: (a) lesions proved at histopathologic examination (n = 4), (b) lesions with characteristic findings at angiography (n = 9) or MR imaging before and after injection of contrast material (n = 38), or (c) lesions with consistent findings at CT and with no change in size for at least 6 months on follow-up radiologic studies (range, 6–42 months; mean, 18 months; n = 58). Angiograms and MR images were available for 11 and 44 lesions, respectively. Angiographic findings were considered to be diagnostic of hemangioma when lesions were seen as round or spherical vascular masses associated with puddling of contrast material. MR findings were considered to be characteristic of hemangioma when lesions showed high signal intensity relative to that of cerebrospinal fluid on T2-weighted images and globular enhancement that progressed centripetally on dynamic contrast material–enhanced images. CT findings were considered to be consistent with hemangioma, as follows: (a) The lesions showed early, peripheral, nodular, broken-ring (or noncontinuous) contrast enhancement; were isoattenuating with the aorta during the HAP; and showed centripetal fill-in enhancement during the PVP (n = 47). (b) Or, the lesions showed early homogeneous enhancement during the HAP and persistent enhancement during the PVP and were isoattenuating with enhanced intrahepatic vessels (n = 11).
CT was performed with spiral scanners (Somatom Plus S, Somatom Plus-4, Siemens Medical Systems, Erlangen, Germany; HiSpeed Advantage, GE Medical Systems, Milwaukee, Wis) in 28, 10, and 31 patients. In all patients, 120 mL of iopromide (Ultravist 370; Schering, Berlin, Germany) was injected into an antecubital vein with a power injector at a rate of 3 mL/sec. HAP and PVP scanning was begun 30 and 65 seconds, respectively, after the start of the injection. A beam collimation of 7 mm and a table speed of 10 mm/sec were used. Images were reconstructed at 7-mm intervals.
MR imaging was performed with a 1.0- (Magnetom Expert; Siemens) or 1.5-T (Magnetom SP, Siemens; Signa, GE Medical Systems) imager. All MR examinations included T2-weighted fast spin-echo imaging (repetition time msec/echo time msec, 3,000–5,000/90–117; echo train length, six to 15), as well as two-dimensional dynamic fast low-angle shot imaging (130–153/5–6, 70° flip angle) performed before then immediately and 1, 3, and 5 minutes after rapid bolus intravenous injection of gadopentetate dimeglumine (Magnevist, Schering; 0.1 mmol per kilogram of body weight).
All radiologic images were retrospectively reviewed by three experienced abdominal radiologists (T.K.K., J.K.H., B.I.C.) by consensus. They evaluated findings on the two-phase spiral CT scans and determined the presence or absence of an arterioportal shunt associated with each hemangioma, the size of the hemangioma, and the rapidity of intratumoral enhancement in each hemangioma. Arterioportal shunts were diagnosed when HAP CT scans showed wedge-shaped or irregularly shaped homogeneous enhancement in the liver parenchyma adjacent to the tumor; when PVP CT scans showed isoattenuation or slight hyperattenuation, compared with adjacent normal liver in that area; and when there was no other demonstrable cause of these attenuation differences, such as segmental portal vein obstruction (11,12). The presence of early opacification of portal branches during the HAP and the attenuation change during the PVP were also evaluated in this wedge-shaped area.
The size of the hemangioma was defined as the greatest diameter on PVP CT images on which the lesion appeared largest. Each hemangioma was categorized as small (<2 cm), medium (2–4 cm), or large (>4 cm).
The rapidity of enhancement was divided into two groups according to the findings on the HAP CT images: rapid when the extent of intratumoral enhancement was more than 50% of the tumor or slow when the extent of intratumoral enhancement was the same or less than 50% of the tumor.
The relationship between the size of the lesion and the presence of an arterioportal shunt was statistically evaluated by means of the Mantel-Haenszel 
2 test, and the rapidity of enhancement of the lesion and the presence of an arterioportal shunt were correlated by means of the Fisher exact test. A P value of less than .05 was considered to indicate a statistically significant difference.
In cases with dynamic contrast-enhanced MR images or hepatic angiograms, these images were also evaluated to determine a wedge-shaped or irregularly shaped homogeneous enhancement in the liver parenchyma (MR images) and early opacification of portal branches (angiograms) adjacent to the tumor.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Among 28 hemangiomas with an arterioportal shunt, the wedge-shaped or irregularly shaped hyperattenuating area adjacent to the tumor during the HAP changed to isoattenuating (Fig 1) in 13 (46%) and slightly hyperattenuating (Fig 2) in 15 (54%) compared with normal liver parenchyma during the PVP. In three lesions (11%), small tubular enhancing structures representing early opacification of portal branches (Fig 2) were seen during the HAP in these areas. MR images and angiograms were available for seven and five lesions, respectively. On dynamic contrast-enhanced MR images, all seven lesions showed transient wedge-shaped or irregularly shaped areas of enhancement, similar to the findings on CT images, in the periphery of the tumors. In all five hemangiomas with hepatic angiograms, angiograms revealed early visualization of portal branches in proximity to the tumors (Fig 3).
|
|
|
|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Cavernous hemangiomas are the most common benign tumor of the liver and have a frequency of up to 7.3% in the general population (17). Because of their histologic characteristics, the majority of hemangiomas show a typical peripheral nodular enhancement with slow progressive centripetal filling on CT scans after bolus injection of contrast material (1–4). However, the enhancement pattern of the hemangiomas can vary. It has been shown (18) that the time necessary for complete isoattenuating fill-in in a hemangioma is related to the internal architecture or collective size of its constituent vascular spaces and differs from tumor to tumor. Therefore, some atypical hemangiomas show immediate homogeneous enhancement and others show minimal or complete lack of enhancement. When the two-phase strategy is used, small hemangiomas with atypical features seemed to have been found more frequently (6,19,20).
An arterioportal shunt is one of the hemodynamic abnormalities that are often detected during the HAP alone. Whereas the arterioportal shunt affecting larger portal branches shows a variable amount of retrograde flow through the portal vein (7), sometimes with reflux to the contralateral side or collaterals in the early phase of dynamic CT (21), that involving smaller portal branches is often demonstrated only as a transient wedge-shaped enhancement adjacent to the tumor (15,16,21–24). While one of the direct CT findings of arterioportal shunt is early demonstration of the affected portal vein within the lesion (21–23), the frequency of this finding is relatively low compared with that at hepatic angiography (15). This is considered to be due to the limited resolution of CT in demonstrating small portal branches of the liver.
Although it is generally recognized that an arterioportal shunt is rare in hepatic hemangiomas (5–10), in our study a relatively high percentage of hemangiomas (25.7%) was accompanied by a wedge-shaped or irregularly shaped parenchymal enhancement adjacent to the tumor at HAP CT; this area usually changed to isoattenuation or slight hyperattenuation, compared with normal liver, during the PVP. In three lesions, HAP CT images showed early depiction of portal branches as well. We believe that these findings most likely represent the existence of arterioportal shunt associated with hepatic hemangioma because the angiograms revealed arterioportal shunts related to the tumors in all five lesions in four patients who underwent hepatic angiography and because there was no other reason for these attenuation differences at two-phase spiral CT.
Whereas portal vein obstruction is considered to be a cause of such attenuation differences in the absence of a arterioportal shunt (11,12), there was no evidence of portal vein thrombosis in our cases. Compression of the portal vein has been uncommonly described (10) only in large hemangiomas. Because hemangiomas are considered to be soft owing to their spongelike histologic characteristics, they are less likely to compress the portal vein and cause such attenuation differences. Therefore, although reports of arterioportal shunts accompanying hepatic hemangiomas have been rare, the frequency of arterioportal shunts associated with hepatic hemangiomas is much greater than previously thought. The rarity of reported cases in the literature may be attributed to the asymptomatic nature of the lesion (25) and to the difficulty in the detection of arterioportal shunts involving small portal branches before introduction of HAP CT.
The results of our study showed that an atypical hemangioma with arterioportal shunt tends to show a rapid enhancement. Similar experiences are documented in prior reports (6,26,27). The pathogenesis of the differences between high- and low-flow hemangiomas has not been fully clarified. However, Yamashita et al (18) suggested that the dynamic enhancement pattern of cavernous hemangioma may correlate with the collective size of its constituent vascular spaces and assumed that tumors with slow fill-in have relatively large vascular spaces and that tumors with rapid enhancement have smaller vascular spaces and large interstitium. Therefore, it might be reasonable to assume that high flow in the smaller intravascular spaces is more likely to produce shunting in the potential communication between the hepatic artery and the portal vein, proved with many clinical and experimental studies (7,22). Most previously reported (6,28) early-enhancing hemangiomas were small (<1 cm). Similarly, most (61%) hemangiomas with arterioportal shunt in our study were small (<2 cm). However, there was no statistically significant relationship between lesion size and presence of the arterioportal shunt.
There are several limitations in our study that deserve mention. First, as is a common limitation with studies of hepatic hemangiomas, our study lacked pathologic correlation with two-phase CT findings. Second, this was a retrospective study and was subject to all of the biases affecting such studies. Third, in most cases, arterioportal shunts were diagnosed exclusively on the basis of CT findings (21–24). Confirmative radiologic diagnosis of an arterioportal shunt often requires celiac or hepatic angiography, which was not performed in most of our cases. However, because angiograms revealed arterioportal shunts related to the tumors in all five lesions for which angiograms were available and because there was no other reason for the attenuation differences on two-phase spiral CT images, we believe that these CT findings support the presence of transtumoral arterioportal shunt in hemangiomas. Finally, our study lacked a comparison with data in population of patients with other hepatic lesions; this may limit the value of our study.
The actual prevalence of arterioportal shunt associated with other hepatic lesions at two-phase spiral CT should be confirmed in another study. However, while previous reports (5,7) describe arterioportal shunts associated with hepatic tumor as a characteristic of malignant tumors, our findings show that benign hepatic hemangiomas with an arterioportal shunt may demonstrate atypical hypervascular enhancement features on HAP CT scans. The major clinical implication of this observation is that because both benign and malignant hepatic tumors may demonstrate an arterioportal shunt, this finding alone does not necessarily indicate that a hepatic tumor is malignant.
In conclusion, arterioportal shunts are not uncommonly seen in hepatic hemangiomas at two-phase spiral CT, and these tumors tend to show rapid enhancement. Therefore, identification of an associated arterioportal shunt on the HAP CT scan does not necessarily imply that the underlying tumor is malignant.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, J.K.H., B.I.C.; study concepts, T.K.K.; study design, K.W.K., T.K.K., J.K.H.; literature research, K.W.K., H.J.L.; clinical studies, K.W.K., H.J.L.; data acquisition, K.W.K., H.J.L., A.Y.K.; data analysis, T.K.K., J.K.H., B.I.C.; statistical analysis, H.J.L.; manuscript preparation, K.W.K.; manuscript definition of intellectual content, T.K.K., B.I.C. manuscript editing, A.Y.K., T.K.K.; manuscript review, A.Y.K, J.K.H.; manuscript final version approval, J.K.H., B.I.C.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. W. Kim, M. J. Kim, S. S. Lee, H. J. Kim, Y. M. Shin, P.-N. Kim, and M.-G. Lee Sparing of Fatty Infiltration Around Focal Hepatic Lesions in Patients with Hepatic Steatosis: Sonographic Appearance with CT and MRI Correlation Am. J. Roentgenol., April 1, 2008; 190(4): 1018 - 1027. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. W. Kim, A. Y. Kim, T. K. Kim, S. Y. Kim, M.-J. Kim, M.-S. Park, S. H. Park, K. H. Lee, J. K. Kim, P.-N. Kim, et al. Hepatic Hemangiomas with Arterioportal Shunt: Sonographic Appearances with CT and MRI Correlation Am. J. Roentgenol., October 1, 2006; 187(4): W406 - W414. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. J. Kim, A. Y. Kim, T. K. Kim, J. H. Byun, H. J. Won, K. W. Kim, Y. M. Shin, P. N. Kim, H. K. Ha, and M.-G. Lee Transient Hepatic Attenuation Differences in Focal Hepatic Lesions: Dynamic CT Features Am. J. Roentgenol., January 1, 2005; 184(1): 83 - 90. [Full Text] [PDF]
|
|
|
|
 |
|
 J. H. Byun, T. K. Kim, C. W. Lee, J. K. Lee, A. Y. Kim, P. N. Kim, H. K. Ha, and M.-G. Lee Arterioportal Shunt: Prevalence in Small Hemangiomas versus That in Hepatocellular Carcinomas 3 cm or Smaller at Two-Phase Helical CT Radiology, August 1, 2004; 232(2): 354 - 360. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H.-J. Jang, T. K. Kim, H. K. Lim, S. J. Park, J. S. Sim, H. Y. Kim, and J.-H. Lee Hepatic Hemangioma: Atypical Appearances on CT, MR Imaging, and Sonography Am. J. Roentgenol., January 1, 2003; 180(1): 135 - 141. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
